There are things that the United States can and should do to revive growth. In broad strokes, I’m for comprehensive tax reform, productivity-enhancing education and health reforms, and a shift in spending priorities from consumption to investment.

But I also think that new forms of wealth are difficult to measure, as economist John Haltiwanger suggests in his paper on “Measuring the Digital Economy.”

The banking sector offers an excellent example of some of the problems confrontingthe statistical agencies due to the IT revolution. The IT revolution has dramatically changedthe nature of banking services with the introduction of electronic banking, ATMs and so on.Statistical agencies have grappled with how to define and measure output in banking for yearsand the IT revolution has simply made this problem worse. For example, it is clear that ATMshave increased the level of service offered by banks to their customers by allowing them toaccess their accounts 24 hours a day and 7 days and week, and by reducing the time theyspend in teller lines. However, the value of these services is not directly measured in anyofficial statistics, but the cost of installing the ATM networks is. Thus, we see that governmentstatistics will understate the productivity increases in banking arising from investments in ITdue to measurement problems.

In Wired for Innovation, Eric Brynjolfsson and Adam Saunders describe how nonmarket transactions enabled by new digital technologies create value without ever being captured in GDP statistics.

In a similar vein, as Will Wilkinson has argued, the fact that we factor the raw amount of government spending into our GDP statistics rather than the value created by effective government spending means that a country like the U.S. with a notoriously inefficient public sector (by the standards of the rich market democracies, at least) might have an undeserved GDP advantage. That is, the inefficiency of our public sector makes us poorer than we would otherwise be while the efficiency of the Swedish public sector might make them richer than they would otherwise be.

So when Neil Irwin describes the output gap in a lucid and intelligent feature for the Washington Post, I can’t help but think about what we’re missing conceptually. What does it mean to have an “output gap” if our understanding of potential GDP relies on dated assumptions regarding what is and what is not valued by consumers? Potential GDP takes, say, the 2007 economy and projects it into the future. But it’s possible that large swathes of the 2007 are now worth much less than they were, and that this step change is permanent. If that is true, we might be chasing the wrong goal.

Rather than try to goose the economy using the crude and ineffective tools at our disposal, we might want to think more seriously about finding ways to employ workers who’ve found their labor market position deteriorate over the last thirty years and, in particular, the last three years.

These thoughts are inspired in part by Lisa Gansky’s excellent new book The Mesh, which she discusses with Dale Dougherty over at O’Reilly Radar. More thoughts on The Mesh to come.